An experimental and CFD study into the dispersion of buoyant gas using passive venting in a small fuel cell enclosure

© 2016 IChemE. With the emergence of a 'Hydrogen Economy', fuel cell (FC) deployment in small enclosures will become common place. However, hydrogen's wide flammable range (4-74%) poses a significant safety concern. Without adequate ventilation, a hydrogen gas leak from a FC could create flammable mixtures in the enclosure, and hence the potential for an explosion. Traditionally, a mechanical ventilation system would be employed in an enclosure to ensure that the hydrogen gas is removed and prevent a flammable concentration forming. However, in many applications (e.g. low power and remote installations) mechanical ventilation is undesirable, since it would drain the FC output and its operation would be vulnerable to any power failures that may occur. In such situations, it is therefore desirable to be able to employ a passive ventilation system to remove the hydrogen gas from the FC enclosure. Passive ventilation relies upon buoyancy driven flow, with the size, shape and position of ventilation openings critical for producing predictable flows and maintaining low gas concentrations. Determining the relationship between gas leak rate, ventilation configuration and internal concentration of buoyant gas will help to inform and optimise FC enclosure safety design. An experimental and Computational Fluid Dynamics (CFD) study was therefore carried out to investigate helium gas dispersion (employed as a safe analogue for hydrogen gas) in a 0.191 m 3 ventilated enclosure. The helium gas was released from a centrally positioned, vertical, 4mm diameter nozzle at low flow rates (1-5 L/min), to simulate a hydrogen leak from a fuel cell (FC) in a small enclosure. A single narrow horizontal vent was created at the top of one vertical face. The helium leak rate was varied and observations of dispersal behaviour and gas concentration made. Similarly positioned vents were introduced on the remaining vertical faces and further observations made. Ventilation flow rates were found to increase as the number of vents increased, and became more effective at keeping helium concentrations below 4% v/v, across the range of leak rates investigated. A cross-flow passive ventilation scheme using opposing lower and upper matched vents provided comparative data. The cross-flow arrangement provided effective displacement ventilation and performed best. The more challenging, high-level vent arrangements provided mixing/exchange ventilation, which became more effective with increasing number of vents. The CFD model was found to be able to replicate the experimental flow behaviour observed, but with variance in concentration levels produced.